GitHub announced generally available Copilot VS Code Agents usage metrics on September 11, 2026. The update adds activity from the dedicated Agents window to enterprise and organization reports. Read GitHub’s announcement.
What changed
For both 1-day and 28-day reports, aggregate data gains optional daily_active_vscode_agent_users and totals_by_vscode_agent fields. User-level reports gain optional used_vscode_agent and totals_by_vscode_agent fields. The totals include session and user-message counts.
GitHub says these measurements cover the dedicated VS Code Agents window separately from editor-window Agent Mode and generic usage rollups. When data is unavailable, optional fields can be absent or null. Access requires an eligible owner, billing-manager or custom role with View Copilot Metrics permission, plus an enabled Copilot usage metrics policy.
Why this matters for QA engineers
QATechTools analysis: This gives teams another adoption signal when piloting agents for test maintenance or failure triage. Session volume alone does not establish that generated tests find defects or that fixes are correct.
Before updating a QA dashboard, test report ingestion with populated, missing and null fields. Preserve “unavailable” as a distinct state so missing telemetry does not become a misleading zero. Label the Agents-window series clearly to avoid mixing different usage surfaces.
For a pilot, compare activity with independently recorded outcomes such as accepted test changes, confirmed defect fixes and review rework. That makes the new reporting useful without treating engagement as a quality score.
News verified September 12, 2026. Source: GitHub Changelog, September 11, 2026.